Here are mightylakers' thoughts on LAS:
Recommendations: 1
I have written a post about LAS-CDMA before on RB. Here it is
What is LAS-CDMA?
LAS stands for the Large Area Synchronous CDMA. Here are the key
prospects about LAS-CDMA
1) As we all know in CDMA, PN codes and Walsh codes are used to
spread the information and differentiate the users. With PN codes used to
identify different cells/sectors and Walsh codes to define different channels
within the same cell/sector. LAS-CDMA claims that that have found different
kind of codes, namely LA codes to replace PN code, and LS code to
replace Walsh code. They claim that LA code and LS code have better
auto-correlation and cross-correlation properties in comparison with PN
code and LS code. Therefore buy using LA codes you can reject cell
interference, by using LS codes you can reject inter-symbol interference
(ISI) and multiple access interference (MAI). Therefore you can greatly
increase the capacity and data rate.
2) Because of the better properties of LA+LS codes. Soft handoff is not
needed because the marginal interference.
3) Using up to 64 QAM(Quadrature Amplitude Modulation) for data
modulation/demodulation to increase the data rate. The peak rate can be
achieved as high as 5.52Mbps
4) The higher layer of LAS-CDMA will remain as the same to be compatible
with CDMA2000. It will operate in the same 1.25Mhz
5) It claims to be compatible with WCDMA and TD-SCDMA too. But that's
just a belief for right now. Nothing has been defined for that part yet.
My thoughts plus some facts:
1) The so-called better properties of LA LS codes are purely theoretical.
They are all based on some conditions that far away from reality. The
so-called interference rejection property is just simply not true. It may or
may not reduce the level of interference. Simulation under some ideal
situation is hardly a proof of anything.
2) More reality checks, the hardest thing in wireless is the signal
transmission and how to deal with multipath, fading, interference. Those
things are hard to understand by just using computer simulation. Without
test it in the real world you can treat those theoretical result onto the shelf.
Solving the problems you encounter in the real world is much more difficult
than the smart guys sitting in the office and write down their imagination. To
put things into perspective. Q proposed the CDMA idea in the mid to late
80's. They conducted the first real CDMA demo call in 91. The first trial
networks in HK was in 95 and till 97 did we start to see the large-scale
commercial networks buildout. So what do you think Q has been doing after
the simulation? It's how to implement the idea into this complex world that
matters. Why do you think Q use what are used today? Those parameters
the result of throughout study of the real world. They are the result of this
optimization process by combining all kinds of characteristics of the real
world. You can always design a simulation and twist the result as will. The
bottom line is we need something works
3) In LAS proposal, it is using 64 QAM to achieve the high data rate. This is
another thing that is far away from reality. By increasing the complexity of
the modulation scheme, you need more computing power and speed. How
can that be supported by today's hardware is a real question. To those of
you knows how modulation works. 64 QAM, in theory, can send 8 bit of
data, 16 QAM 4 and 8-QPSK for 3 bits. Currently HDR is using 8-QPSK or
16QAM. And by theory using 64QAM will increase the speed into 4.8Mbps
from 2.4Mbps. But the complexity of the system will increase multiple fold.
Is the cost worth of it? And again don't forget the reality. I'm basing HDR
argument on something that already works in the real world, not some
simulation stuffs under ideal condition. One thing in commercial product is
not only the quality of the product, but also the cost are very important too.
There are always tradeoffs between those two. Q is not hyping up the
crowd by claiming much higher data rate based on some theoretical
modulation scheme. You may ask what if the hardware is good enough to
do it. I say then let's do it that way by simply upgrade that part and here we
go.
4) About the recent LAS-CDMA voice call being conducted. We have no
info as for how the test is done with what kind of equipment under what
kind of condition. It wasn't even a demo. Overall it is still a very premature
technology(theory) with more questions than answers. In fact as you can
see even with our current working system we can jack up the numbers that
are close to LAS claimed numbers by just using ideal assumption. So even
if by a snowball chance that all things claimed by LAS are true. The
question is that is it worth of such a total revamp to achieve just a very
marginal improvement? And don't forget after all, it is still using a lot of
CDMA2000 stuffs. The main thing is just what kind of codes to use. So
guess what, I think you all know it:-) |
